I'm of the firm belief that two weeks is a magical number.
The human brain has a few factors working against it when you're trying to do something new, learn something, or accomplish some feat you have in mind.
Attrition works against you (every day's attentions can be different). Your peaks and lows for any given day can work for or against you.
Two weeks will give you enough time to do the following cycle:
Validate, mull over the idea of your goal in your mind.
Try towards it and fail a few times. Revamp the idea.
Step away and come back a few times. Have some tea.
Reach a straining point or natural wall: "I can't go any further".
Important: you get to choose if it's important enough to proceed beyond that natural wall. Stop here if not.
You've identified an edge you have over others. Rinse and repeat.
Most people never proceed from 5 to 6. In two weeks of time, at 6 hours a day, you have 5000 minutes of attention spent on something that you can decide if it's worth pursuing or not. In the age of AI, where it's easier than ever to learn and change your direction, understanding true center is the game changer.
Most people spend their entire lives trying to find their true center - their passions, motivations, directions, etcetera; I argue you can triangulate it in weeks (especially today, with the tools available to you).
I'll give an example: in one of my past articles, I described designing actual circuit boards for my own personal use cases. I'm still debugging the actual traces/design (I'm not an electrical engineer by trade), but two weeks of concerted effort gave me enough time to wonder: "is this direction of capturing idiosyncratic audio worth it?"
Sure, maybe I'm not of a traditional background there, but I am able to design towards my direction, even if it's fraught with stumbles. Nikita Bier, now head of product at X, has this quote (which is a favorite of mine):
Still working on circuitry; tscircuit lets me design actual PCBs using AI and TypeScript, which is nuts (we live in the future), but I discovered that on top of caring about the actual data (i.e. audio), I've started caring about the full end-to-end product (from atomic construction all the way to digital programming/representation).
In today's world of AI, I must understand and own the full stack. With robotics coming into play (Unitree is fascinating), this physical hardware angle gives me directional correctness.
You cannot go wrong with concerted efforts to align your own macro direction - and it starts with giving dedicated time to things. Two weeks should be a good starting point, I'd wager.
Be well,
Michael Kirsanov